Sage Journals: Discover world-class research

Abstract

Introduction

Immune checkpoint inhibitors plus chemotherapy has become the new standard of care for advanced biliary tract cancer (BTC). However, the real-world effectiveness and safety of this approach, and its association with molecular alterations, remains uncertain.

Methods

In this single institutional retrospective cohort study, patients with advanced BTC treated with systemic chemotherapy with or without immunotherapy from July 2020 to June 2025 were included. Primary endpoint was overall survival (OS) in patients treated with first-line chemoimmunotherapy; secondary endpoints included progression-free survival (PFS), objective response rate (ORR), safety, prognostic factors, and molecular profiling.

Results

49 of 52 patients (94.2%) received chemoimmunotherapy. Median OS and PFS were 10.2 months and 5.4 months, respectively; the ORR was 24.5%. Poor Eastern Cooperative Oncology Group (ECOG) performance status and hypoalbuminemia were independent predictors of poor OS. Among the 52 patients, molecular profiling identified IDH1 mutations (9.6%), FGFR2 fusions (3.9%), HER2 amplification (5.8%), dMMR (5.8%), and BRCA mutations (5.8%). HER2 amplification was associated with worse OS; dMMR and BRCA mutations showed a trend toward improved OS.

Conclusion

Real world first-line chemoimmunotherapy for advanced BTC achieved outcomes comparable to those in pivotal trials. The ECOG performance status, serum albumin level, and HER2 status were prognostic for OS, supporting the importance of baseline patient selection and molecular profiling in treatment planning.

Keywords

cholangiocarcinoma gallbladder cancer chemotherapy immunotherapy efficacy biomarkers IDH1 FGFR2 HER2 prognostic factors

Introduction

Biliary tract cancers (BTCs), comprising intrahepatic cholangiocarcinoma (iCCA), extrahepatic cholangiocarcinoma (eCCA), and gallbladder carcinoma (GBC), are aggressive malignancies with poor prognoses and limited therapeutic options.^1-3 Although rare in Western countries, their incidence has been increasing worldwide, particularly iCCA.^4,5 Most patients present with advanced-stage disease, for which systemic chemotherapy remains the mainstay of treatment. However, outcomes remain suboptimal, with median overall survival (OS) rarely exceeding 1 year.^6-8

For over a decade, gemcitabine combined with cisplatin has been the standard first-line therapy for advanced BTC.⁶ Recent phase III trials, including TOPAZ-1 and KEYNOTE-966, have demonstrated the clinical benefit of adding immune checkpoint inhibitors to standard chemotherapy in the first-line setting.^7,8 These studies showed modest but statistically significant improvement in survival with durvalumab and pembrolizumab, respectively, leading to the establishment of chemotherapy plus immunotherapy as the new standard of care for advanced BTC.

Concurrently, the increasing use of molecular profiling has identified a spectrum of potentially targetable genomic alterations in BTC, including fibroblast growth factor receptor 2 (FGFR2) fusions, isocitrate dehydrogenase 1 (IDH1) and B-Raf (BRAF) mutations, and human epidermal growth factor receptor 2 (HER2) amplification.^5,9-12 These biomarkers may be prognostic and are of increasing predictive significance,^13-19 although their clinical utility in the context of chemoimmunotherapy remains to be fully defined. Despite the paradigm shift in treatment, evidence for the real-world effectiveness of chemotherapy combined with immunotherapy remains scarce. Clinical trials typically enroll highly selected patients under controlled conditions, and their findings may not fully reflect outcomes in routine oncology practice.

This retrospective study aimed to evaluate the real-world effectiveness of chemotherapy combined with immunotherapy in patients with advanced BTC and to describe the molecular alteration landscape in this population. Our findings may complement clinical trial evidence, support decision-making in clinical practice, and provide insights into the potential role of molecular profiling in the contemporary management of advanced BTC.

Materials and Methods

Study Design and Participants

This was a retrospective cohort study conducted at the Peter MacCallum Cancer Centre, Melbourne, Australia. Medical records of patients diagnosed with BTC between July 2020 and June 2025 were reviewed. Eligible patients were required to meet the following criteria: (1) histologically confirmed iCCA, eCCA, or GBC; (2) advanced disease at the time of treatment, defined as unresectable or metastatic; (3) receipt of first-line palliative systemic chemotherapy (gemcitabine plus cisplatin or carboplatin) with or without immune checkpoint inhibitors (durvalumab or pembrolizumab); and (4) age ≥18 years. Patients were excluded if they had (1) histological subtypes other than adenocarcinoma, or (2) received only best supportive care without systemic treatment. Clinical information was extracted from the institution’s electronic medical record system. All eligible patients were included in this study to provide a real-word perspective.

The study protocol was approved by the Human Research Ethics Committee and Governance Office of the Peter MacCallum Cancer Centre (HREC: QA/121051/PMCC/September 3, 2025). Owing to the retrospective study design, the requirement for written informed consent was waived. All data were de-identified to maintain patient confidentiality and adhere to privacy regulations. Patients were enrolled until June 2025, and data were censored at the institutional cut-off (July 30, 2025). The reporting of this study conformed to STROBE guidelines.²⁰

Outcome Measures

The primary outcome was OS in patients with advanced BTC who received first-line palliative chemotherapy in combination with immunotherapy. Secondary outcomes included progression-free survival (PFS), objective response rate (ORR), safety, and identification of prognostic factors for OS in this treatment setting. Additionally, biomarker analyses from all available patients to characterize molecular alterations were conducted as exploratory analyses.

OS was defined as the time from the initiation of systemic therapy to death from any cause. PFS was measured from the start of systemic therapy until either radiologically confirmed disease progression or death, whichever occurred first. Tumor assessments were performed every 2-3 months using chest and abdominal computed tomography. Treatment response was evaluated in accordance with the Response Evaluation Criteria in Solid Tumors version 1.1 using official institutional radiology reports; no retrospective central re-review of imaging was undertaken. Adverse events were graded by the treating oncologists according to the Common Terminology Criteria for Adverse Events version 5.0, using clinical, biochemical, and radiographic features consistent with the observed toxicities.

The treatment regimen consisted of durvalumab or pembrolizumab in combination with gemcitabine and either cisplatin or carboplatin, administered intravenously in 21-day cycles for up to 8 cycles. Durvalumab was administered at 1500 mg on day 1 of each cycle, in combination with gemcitabine (1000 mg/m²) and cisplatin (25 mg/m²) on days 1 and 8. After completion of the chemotherapy, durvalumab monotherapy (1500 mg every 4 weeks) was continued until disease progression, unacceptable toxicity, patient withdrawal, or death. When cisplatin was substituted with carboplatin, it was administered at an AUC of 1.5 on days 1 and 8. For patients receiving pembrolizumab instead of durvalumab, the dose was 200 mg on day 1 of each cycle, with administration every 3 weeks thereafter. Treatment selection was determined by the treating oncologists in agreement with the patients, based on baseline characteristics, drug accessibility, and reimbursement policies during each period. This could introduce potential selection bias in both treatment allocation and molecular testing. However, the primary aim of this study was to describe real-world outcomes.

Statistical Analysis

Baseline demographic and clinical characteristics were summarized using descriptive statistics. Continuous variables are presented as mean ± standard deviation (SD) for normally distributed data, or median with interquartile range (IQR) for non-normally distributed data. Categorical variables are reported as absolute frequencies and percentages. Survival outcomes, including OS and PFS, were estimated using the Kaplan–Meier method with censoring applied for patients alive at the data cut-off, and median survival times with corresponding 95% confidence intervals (CIs) were reported. Differences between survival curves were assessed using the log-rank test. Univariable Cox proportional hazards regression was used to assess factors associated with OS, and variables deemed potentially significant (P < 0.10 in univariable analysis) were subsequently included in a multivariable Cox regression to identify independent prognostic factors, reported as hazard ratios (HRs) with 95% CIs. The proportional hazards assumption was evaluated using Schoenfeld residuals. Model stability was assessed by the events-per-variable (EPV) criterion for the final multivariable model, with EPV ≥10 considered acceptable. Analyses were conducted on an available-case basis without imputation. All statistical analyses were conducted using R software, version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). A two-sided P-value of <0.05 was considered statistically significant.

Results

Baseline Characteristics

A total of 52 patients were included in this study, with 49 (94.2%) receiving first-line chemotherapy plus immunotherapy (Table 1). The median age was 64.7 years (IQR, 53.6-69.3), and 23.1% were aged ≥70 years. Most patients had good Eastern Cooperative Oncology Group (ECOG) performance status (PS), while 15.4% had an ECOG PS of 2. iCCA was the most common primary site (55.8%), followed by GBC (23.1%) and eCCA (21.2%). The majority (90.4%) presented with metastatic disease, predominantly as de novo metastases (76.9%). The liver (57.7%) and distant lymph nodes (53.8%) were the most frequent metastatic sites, with more than half of patients having metastases at 2 or more sites.

Table 1.

Baseline Characteristics of the Patients

Characteristics	Total (n = 52)	First-line chemotherapy plus immunotherapy (n = 49)
Sex, n (%)
Female	27 (51.9)	25 (51.0)
Male	25 (48.1)	24 (49.0)
Age, years (IQR)	64.7 (53.6-69.3)	64.7 (55.8-69.5)
Age ≥70 years, n (%)	12 (23.1)	12 (24.5)
BMI, n (%)
<18.5 kg/m²	1 (1.9)	1 (2.0)
18.5-24.9 kg/m²	23 (44.2)	22 (44.9)
25-29.9 kg/m²	17 (32.7)	16 (32.7)
≥30 kg/m²	11 (21.2)	10 (20.4)
ECOG PS, n (%)
0	27 (51.9)	24 (49.0)
1	17 (32.7)	17 (34.7)
2	8 (15.4)	8 (16.3)
Comorbidities, n (%)
Coronary artery disease	4 (7.7)	4 (8.2)
Chronic kidney disease	2 (3.8)	2 (4.1)
Cerebrovascular disease	1 (1.9)	1 (2.0)
Cirrhosis	0 (0)	0 (0)
Viral hepatitis B infection	0 (0)	0 (0)
Viral hepatitis C infection	0 (0)	0 (0)
Location, n (%)
iCCA	29 (55.8)	27 (55.1)
eCCA	11 (21.2)	11 (22.4)
GCB	12 (23.1)	11 (22.4)
Advanced stage, n (%)
Metastatic disease	47 (90.4)	44 (89.8)
De novo disease	40 (76.9)	39 (79.6)
Recurrent disease	7 (13.5)	5 (10.2)
Unresectable disease	5 (9.6)	5 (10.2)
Number of site metastasis, n (%)
0	5 (9.6)	5 (10.2)
1	19 (36.5)	18 (36.7)
≥2	28 (53.9)	26 (53.1)
Site of metastasis, n (%)
Liver	30 (57.7)	28 (57.1)
Distant lymph node	28 (53.8)	26 (53.1)
Lung	14 (26.9)	13 (26.5)
Peritoneum	8 (15.4)	8 (16.3)
Bone	7 (13.5)	7 (14.3)
Others	1 (1.9)	1 (2.0)
Differentiation, n (%)
Well	3 (5.8)	3 (6.1)
Moderately	19 (36.5)	16 (32.7)
Poorly	30 (57.7)	30 (61.2)
Albumin, g/dL (IQR)	3.3 (3.1-3.8)	3.2 (3.0-3.7)
Albumin <3.5 g/dL, n (%)	29 (55.8)	29 (59.2)
CA19-9, U/mL (IQR)	235.5 (18.0-3556.5)	297.0 (19.0-3759.0)
CA19-9 ≥ 37, n(%)	33 (63.5)	32 (65.3)
First-line systemic treatment, n (%)
Chemotherapy plus immunotherapy	49 (94.2)	49 (100.0)
Cisplatin plus gemcitabine and durvalumab	42 (80.8)	42 (85.7)
Cisplatin plus gemcitabine and pembrolizumab	4 (7.7)	4 (8.2)
Carboplatin plus gemcitabine and durvalumab	3 (5.8)	3 (6.1)
Chemotherapy alone	3 (5.8)	-
Cisplatin plus gemcitabine	2 (3.9)	-
Gemcitabine plus tegafur	1 (1.9)	-

BMI, body mass index; ECOG, Eastern Cooperative Oncology Group; PS, performance status; CA19-9, cancer antigen 19-9; IQR, interquartile range; iCCA, intrahepatic cholangiocarcinoma; eCCA, extrahepatic cholangiocarcinoma; GBC, gallbladder carcinoma.

Biomarker Assessment

Molecular profiling results based on approved targetable treatments are summarized in Table 2. Mismatch repair (MMR) status was assessed in 90.4% of patients, HER2 overexpression/amplification in 34.6%, and next-generation sequencing (NGS) conducted in 55.8%. Among those undergoing NGS, 65.5% were tested using TruSight Oncology 500 and 34.5% with the Oncomine platform.

Table 2.

Biomarker Assessment (N = 52)

Biomarker	N (%)
MMR status
Proficient	44 (84.6)
Deficient	3 (5.8)
Not assessed	5 (9.6)
HER2 overexpression/amplification
Positive	3 (5.8)
IHC 3+	2 (3.9)
IHC 2+/FISH amplified	1 (1.9)
Negative	15 (28.9)
IHC 0	5 (9.6)
IHC 1+	3 (5.8)
IHC 2+/FISH not amplified	7 (13.5)
Not assessed	34 (65.4)
PD-L1
CPS ≥1	4 (7.6)
CPS 0	2 (3.9)
Not assessed	46 (88.5)
NGS
Tested	29 (55.8)
HER2 gain	2 (3.9)
IDH1 mutation	5 (9.6)
FGFR2 fusion	2 (3.9)
NTRK Fusion	0 (0)
TMB-high	1 (1.9)
BRAF V600E mutation	0 (0)
RET Fusion	0 (0)
KRAS G12C mutation	0 (0)
BRCA Mutation	3 (5.8)
Not tested	23 (44.2)

BRAF, b-Raf; BRCA, BReast CAncer gene; CPS, combined positive score; FGFR2, fibroblast growth factor receptor 2; FISH, fluorescence in situ hybridization; HER2, human epidermal growth factor receptor 2; IDH1, isocitrate dehydrogenase 1; IHC, immunohistochemical testing; KRAS, K-Ras; MMR, mismatch repair; NGS, next-generation sequencing; NTRK, neurotrophic tyrosine receptor kinase; PD-L1, programmed death-ligand 1; TMB, tumor mutational burden.

Of all 52 patients, deficient mismatch repair (dMMR) was identified in 5.8%, and HER2 overexpression/amplification in 5.8%. NGS revealed targetable molecular alterations including IDH1 mutation (9.6%), FGFR2 fusion (3.9%), and high tumor mutational burden (TMB-high, 1.9%). No cases were identified with neurotrophic tyrosine receptor kinase (NTRK) fusions, BRAF V600E mutations, K-Ras (KRAS) G12 C mutations, or RET fusions. Baseline characteristics of patients by molecular alterations is presented in Supplemental Table S1. The complete spectrum of molecular alterations detected by NGS is shown in Figure 1 and detailed in Supplemental Table S2.

Figure 1.

Molecular Alterations Findings From Next-Generation Sequencing Testing (N = 29)

Real-World Effectiveness of First-Line Chemotherapy Plus Immunotherapy

A total of 49 patients treated with first-line chemotherapy plus immunotherapy were included in the effectiveness analysis, which evaluated OS, PFS, ORR, safety, and prognostic factors for OS.

Treatment Information

Of the 49 patients, 42 (85.7%) received cisplatin plus gemcitabine with durvalumab; 4 (8.2%) received carboplatin plus gemcitabine with durvalumab; and 3 (6.1%) received cisplatin plus gemcitabine with pembrolizumab. Treatment details are summarized in Table 3. The most common reason for chemotherapy discontinuation was disease progression (34.7%), followed by treatment completion (26.5%) and declining PS (18.4%). Immunotherapy was most frequently discontinued owing to disease progression (57.1%), declining PS (22.4%), and immune-related adverse events (irAEs) (11.9%). At the data cut-off, 93.9% of patients had discontinued first-line chemoimmunotherapy. Second-line treatments were administered to 38.8% of patients, predominantly chemotherapy (22.4% FOLFOX, 8.2% FOLFIRI) and targeted therapies, including futibatinib, ivosidenib, olaparib, and pamiparib plus tislelizumab (as part of a clinical trial). A few patients received third-line or later therapies. Details of the 3 patients treated with chemotherapy alone are provided in Supplemental Table S3.

Table 3.

Treatment Information for Patients Who Received First-Line Chemotherapy Plus Immunotherapy (N = 49)

Variable	Value
Number of cycles of chemotherapy, median (IQR)	5 (2,8)
Number of cycles of immunotherapy, median (IQR)	5 (2,9)
Any chemotherapy dose reduction, n (%)	34 (69.4)
Discontinuation of chemotherapy, n (%)
Disease progression	17 (34.7)
Treatment completion	13 (26.5)
Decline in ECOG PS	9 (18.4)
Severe infection/Adverse events	6 (12.2)
Patient preference	1 (2.0)
Death	1 (2.0)
Transfer of care	2 (4.0)
Discontinuation of immunotherapy, n (%)
Disease progression	28 (57.1)
Decline in ECOG PS	11 (22.4)
Immune-related adverse events	5 (11.9)
Transfer of care	2 (4.0)
Ongoing treatment	3 (6.1)
Second-line therapy, n (%)	19 (38.8)
Chemotherapy
FOLFOX	11 (22.4)
FOLFIRI	4 (8.2)
Targeted therapy
Futibatinib	1 (2.0)
Ivosidenib	1 (2.0)
Olaparib	1 (2.0)
Pamiparib plus tislelizumab	1 (2.0)
Third-line therapy, n (%)	4 (8.2)
Chemotherapy
FOLFIRI	2 (4.1)
Paclitaxel	1 (2.0)
Targeted therapy
Ivosidenib	1 (2.0)
Fourth-line therapy, n (%)	1 (2.0)
Chemotherapy
FOLFIRI	1 (2.0)

FOLFOX, 5-fluorouracil plus leucovorin and oxaliplatin; FOLFIRI, 5-fluorouracil plus leucovorin and irinotecan; ECOG, Eastern Cooperative Oncology Group; PS, performance status.

OS

The median follow-up duration was 8.4 months (range, 0.5-53.5 months). At the data cut-off, 35 patients (71.4%) had died. The median OS for the entire cohort was 10.2 months (95% CI, 6.7-15.3), with 1-year and 2-year OS rates of 35.6% and 11.9%, respectively (Figure 2A).

Figure 2.

Overall Survival (A) and Progression-free Survival (B) in Patients Who Received First-Line Chemotherapy Plus Immunotherapy

When stratified by first-line regimen, the median OS was 10.2 months for cisplatin plus gemcitabine with durvalumab, 7.4 months for carboplatin plus gemcitabine with durvalumab (HR, 1.05; 95% CI, 0.31-3.60; P = 0.940), and 14.3 months for cisplatin plus gemcitabine with pembrolizumab (HR, 0.56; 95% CI, 0.17-1.87; P = 0.347) (Supplemental Figure S1).

When stratified by primary tumor location, the median OS was 10.6 months for iCCA, 10.2 months for eCCA (HR, 1.20; 95% CI, 0.52-2.76; P = 0.673), and 3.7 months for GBC (HR, 1.83; 95% CI, 0.81-4.14; P = 0.148) (Supplemental Figure S2).

The PS at baseline was strongly associated with survival. Patients with ECOG PS 0 had a median OS of 15.5 months, compared with 6.0 months for ECOG PS 1 (HR, 4.06; 95% CI, 1.83-8.99; P < 0.001) and 2.2 months for ECOG PS 2 (HR, 22.32; 95% CI, 7.23-68.90; P < 0.001) (Supplemental Figure S3).

PFS

The median PFS for the overall cohort was 5.4 months (95% CI, 3.8-8.4; Figure 2B). When stratified by first-line regimen, the median PFS was 5.4 months for cisplatin plus gemcitabine with durvalumab, 3.0 months for carboplatin plus gemcitabine with durvalumab (HR, 2.01; 95% CI, 0.60-6.70; P = 0.255), and 6.9 months for cisplatin plus gemcitabine with pembrolizumab (HR, 0.56; 95% CI, 0.17-1.86; P = 0.342) (Supplemental Figure S4).

When stratified by primary tumor location, the median PFS was 5.8 months for iCCA, 4.4 months for eCCA (HR, 0.91; 95% CI, 0.42-1.98; P = 0.813), and 3.7 months for GBC (HR, 1.50; 95% CI, 0.70-3.21; P = 0.293) (Supplemental Figure S5).

Baseline ECOG PS was strongly associated with PFS. Median PFS was 9.7 months for ECOG PS 0, 3.8 months for ECOG PS 1 (HR, 4.47; 95% CI, 2.08-9.63; P < 0.001), and 1.7 months for ECOG PS 2 (HR, 19.51; 95% CI, 6.56-58.06; P < 0.001) (Supplemental Figure S6).

Response rates

The best overall response to first-line chemotherapy plus immunotherapy is summarized in Table 4. Response rates stratified by treatment regimen and by primary tumor location are provided in Supplemental Tables S4 and S5.

Table 4.

Best Response to First-Line Chemotherapy Plus Immunotherapy

Response	n (%)
Complete response	0 (0)
Partial response	12 (24.5)
Stable disease	18 (36.7)
Progressive disease	7 (14.3)
Not assessed	12 (24.5)
Objective response rate	12 (24.5)

Safety

Among patients receiving chemotherapy plus immunotherapy (N = 49), the most frequently reported adverse events of any grade were anemia (75.5%), fatigue (73.5%), and elevated alanine aminotransferase (ALT) levels (61.2%) (Table 5). The most common grade 3-4 toxicities were neutropenia (32.7%), infection (28.6%), and anemia (16.3%). irAEs occurred in 28.6% of patients. Most irAEs, including rash, arthritis, hyperthyroidism, hypothyroidism, and hypophysitis, were grades 1-2. However, severe (grades 3-4) immune-related hepatitis and colitis were observed in 8.2% and 4.1% of patients, respectively. irAEs were managed according to institutional practice consistent with international guideline recommendations.

Table 5.

Adverse Events in Patients Receiving First-Line Chemotherapy Plus Immunotherapy (N = 49)

Adverse events, n (%)	All grades (n = 49)	Grades 1-2 (n = 49)	Grades 3-4 (n = 49)
Hematological
Neutropenia	26 (53.1)	10 (20.4)	16 (32.7)
Anemia	37 (75.5)	29 (59.2)	8 (16.3)
Thrombocytopenia	25 (51.0)	23 (46.9)	2 (4.1)
Febrile neutropenia	5 (10.2)	-	5 (10.2)
Non-hematological
Nausea	12 (24.4)	11 (22.4)	1 (2.0)
Vomiting	2 (4.1)	1 (2.0)	1 (2.0)
Diarrhea	6 (12.3)	4 (8.2)	2 (4.1)
Peripheral neuropathy	11 (22.4)	11 (22.4)	0 (0)
Infection	14 (28.6)	0 (0)	14 (28.6)
Fatigue	36 (73.5)	32 (65.3)	4 (8.2)
Elevated aspartate aminotransferase	21 (42.9)	17 (34.7)	4 (8.2)
Elevated alanine aminotransferase	30 (61.2)	25 (51.0)	5 (10.2)
Immune-related adverse events	14 (28.6)	8 (16.3)	6 (12.3)
Immune-related hepatitis	5 (10.2)	1 (2.0)	4 (8.2)
Immune-related colitis	2 (4.1)	0 (0)	2 (4.1)
Immune-related hyperthyroidism	2 (4.1)	2 (4.1)	0 (0)
Immune-related hypothyroidism	2 (4.1)	2 (4.1)	0 (0)
Immune-related hypophysitis	2 (4.1)	2 (4.1)	0 (0)
Immune-related rash	6 (12.2)	5 (10.2)	1 (2.0)
Immune-related arthritis	2 (4.1)	2 (4.1)	0 (0)

Prognostic factors for OS

In multivariable Cox regression analysis, ECOG PS and baseline serum albumin level were independently associated with OS (Table 6). Compared with ECOG PS 0, both ECOG PS 1 (HR, 2.98; 95% CI, 1.41-6.30; P = 0.004) and ECOG PS 2 (HR, 13.11; 95% CI, 4.93-34.81; P < 0.001) were associated with significantly worse survival. Hypoalbuminemia (<3.5 g/dL) was also a strong predictor of poor prognosis (HR, 11.37; 95% CI, 3.13-41.26; P < 0.001). The proportional hazards assumption was not violated (Schoenfeld residuals: global P = 0.67; ECOG P = 0.88; albumin P = 0.33; CA19-9 P = 0.55).

Table 6.

Prognostic Factors for OS

Factors	Univariate analysis		Multivariate analysis
Factors	HR (95% CI)	P value	HR (95% CI)	P value
Sex: Male vs Female	1.34 (0.68-2.64)	0.394	-	-
Age ≥70 years	0.85 (0.40-1.83)	0.683	-	-
BMI (kg/m²)
<18.5	Ref		-	-
18.5-24.9	0.21 (0.03-1.73)	0.148
25.0-29.9	0.25 (0.03-2.11)	0.203
≥30	0.17 (0.02-1.52)	0.112
ECOG PS
0	Ref		Ref
1	4.06 (1.83-8.99)	<0.001	3.21 (1.35-7.63)	0.008
2	22.32 (7.23-68.90)	<0.001	12.39 (3.78-40.61)	<0.001
Albumin <3.5 g/dL	10.64 (3.95-28.67)	<0.001	8.32 (2.65-26.07)	<0.001
CA19-9 ≥ 37 U/mL	2.47 (1.09-5.58)	0.030	1.07 (0.44-2.56)	0.888
Location
Intrahepatic CCA	Ref		-	-
Extrahepatic CCA	1.20 (0.52-2.76)	0.673
Gallbladder carcinoma	1.83 (0.81-4.14)	0.148
Differentiation
Well	Ref		-	-
Moderately	2.16 (0.27-17.24)	0.466
Poorly	6.27 (0.83-47.09)	0.074
Advanced-stage disease
Locally advanced unresectable	Ref		-	-
Metastatic	1.24 (0.43-3.55)	0.687
Number of metastatic sites
0	Ref		-	-
1	1.40 (0.45-4.33)	0.563
≥2	1.15 (0.38-3.43)	0.808

BMI, body mass index; ECOG, Eastern Cooperative Oncology Group; PS, performance status; CA19-9, cancer antigen 19-9; CCA, cholangiocarcinoma; HR, hazard ratio; CI, confidence interval.

An exploratory analysis was performed to assess the prognostic impact of molecular alterations (Supplemental Table S6). HER2 overexpression/amplification was associated with markedly worse OS (HR, 14.09; 95% CI, 2.31-85.74; P = 0.004), whereas dMMR and BRCA mutation showed a trend toward improved OS, although not statistically significant. For FGFR2 fusion and TMB-high, only 1 patient was present in each subgroup, both of whom were alive at the data cut-off, precluding HR estimation. A swimmer plot demonstrating molecular alteration in each patient, and time to progression and death, is presented in Figure 3.

Figure 3.

A Swimmer Plot Demonstrating Time to Disease Progression and Death for Patients Who Received Chemotherapy Plus Immunotherapy. Each bar Represents One Patient in the Study and Their Length of Survival in Months (N = 49)

Discussion

This real-world study demonstrated the effectiveness of first-line chemotherapy combined with immunotherapy in patients with BTC. In our cohort, the median OS was 10.2 months, median PFS was 5.4 months, and ORR was 24.5%. These outcomes are broadly comparable to those reported in pivotal phase III trials such as TOPAZ-1 (median OS 12.8 months, PFS 7.2 months) and KEYNOTE-966 (median OS 12.7 months, PFS 6.5 months),^7,8 although both OS and PFS were slightly lower in our cohort. The most likely explanation lies in the differences in patient selection. Our study included a real-world population with 15.4% ECOG PS 2, a group typically excluded from clinical trials. Consistent with this, survival outcomes differed markedly by baseline PS, with a median OS of 15.5, 6.0, and 2.2 months and median PFS of 9.7, 3.8, and 1.7 months for ECOG PS 0, 1, and 2, respectively. These results emphasize that while chemoimmunotherapy is effective, its benefit is maximized in patients with good baseline functional status.

Our findings are in line with a real-world Taiwanese study series (median PFS 5.6 months) that also included ECOG PS 2 patients,²¹ but their longer OS (median OS 15.8 months) may be attributable to younger age, more locally advanced rather than metastatic disease, and a lower proportion of GBC, which is generally associated with poorer outcomes.^6,22,23

Another factor potentially influencing OS is the lower rate of subsequent systemic therapy in our study (38.8%) compared to that in TOPAZ-1 (54.8%). In TOPAZ-1, median OS was 16.6 months in patients who received subsequent treatment vs 7.9 months in those who did not, underscoring the prognostic value of post-progression therapy.⁷ The shorter OS in our cohort may partly reflect this gap in access to second-line options. Ensuring timely access to effective subsequent therapies could extend survival in the BTC population.

Baseline ECOG PS and serum albumin level were independent prognostic factors in our cohort, confirming prior studies that identified poor functional status and hypoalbuminemia as markers of inferior survival in BTC.^24-28 While cancer antigen 19-9 level was not independently associated with OS in our multivariable model, it showed significance in univariate analysis, suggesting potential prognostic relevance depending on the cutoff thresholds used in future studies. Altogether, baseline functional and nutritional assessments should be integral to treatment planning in BTC.

The ORR of 24.5% in our cohort was consistent with the 26%-29% reported in the pivotal trials,^7,8,29 despite nearly one-quarter of patients lacking a response evaluation. A response was observed across all primary tumor sites, although GBC maintained a poorer survival prognosis despite reported higher response rates in some earlier studies.^6,30,31 These findings reinforce the aggressive biology of GBC and the need for site-specific research, especially focusing on molecular alterations. In our cohort, patients with ECOG PS 2 and those with GBC appeared to have the poorest OS and PFS. However, these findings should be interpreted with caution given the small sample size and single-center setting. While chemoimmunotherapy may still be a treatment option in this subgroup, it should be considered with careful discussion, close monitoring, and efforts to optimize functional status.

The safety profile was generally consistent with clinical trial data, with anemia, neutropenia, thrombocytopenia, elevated aspartate aminotransferase/ALT levels, nausea, and fatigue being the most common adverse events. irAEs occurred in 28.6% of patients, with 12.3% being grades 3-4, and leading to discontinuation of immunotherapy in 11.9%. This rate is higher than the 2%-7% reported in TOPAZ-1 and KEYNOTE-966,^7,8,29 with hepatitis being the most frequent severe irAE. Overall, safety profile is manageable but vigilance for irAEs is crucial, reinforcing the need for close monitoring and rapid management of irAEs.

Molecular profiling, performed in 55.8% of patients, identified potentially targetable alterations: IDH1 mutations (9.6%), FGFR2 fusions (3.9%), dMMR (5.8%), TMB-high (1.9%), HER2 amplification (5.8%), and BRCA mutations (5.8%). When restricted to patients tested using NGS, IDH1 and FGFR2 alterations were found in 17.2% and 6.9%, respectively, suggesting an underestimation because of incomplete testing. This incomplete testing rate reflects real-world practice, where tissue availability and test accessibility are variable, and may introduce selection bias in biomarker prevalence estimates. These findings are consistent with previous reports for IDH1 mutations (10%-20%) but slightly lower for FGFR2 fusions (10%-15%) and HER2 amplification (typically 10%-20% in eCCA and GBC).^11,32,33 IDH1 mutation and FGFR2 fusion were observed almost exclusively in iCCA, whereas HER2 overexpression/amplification occurred mainly in eCCA and GBC. Although the prevalence of actionable alterations is relatively low, comprehensive molecular profiling, preferably using NGS to cover the full spectrum of targets,¹⁰ is essential to identify patients who could benefit from targeted therapy.

In exploratory biomarker analyses, HER2 overexpression/amplification was associated with significantly poorer OS, whereas dMMR and BRCA mutations showed a non-significant trend toward better survival. These findings align with previous studies reporting HER2 as a negative prognostic marker in BTC³⁴ and support the integration of comprehensive molecular testing into standard care. For HER2-positive BTC, evidence from gastric cancer suggests potential benefit from HER2-targeted therapy in combination with chemotherapy and immunotherapy,³⁵ an approach that warrants evaluation in prospective BTC trials. Taken together, routine comprehensive genomic testing can identify high-risk subgroups and guide targeted therapy strategies. Because molecular testing was not universally performed, biomarker missingness may be non-random and could bias alteration prevalence and reduce precision of exploratory subgroup analyses; accordingly, these findings should be interpreted with caution and results should be viewed as hypothesis-generating.

Although no predictive biomarkers have been validated for chemoimmunotherapy benefit in BTC, molecular testing remains essential to guide second-line therapy choices, particularly for IDH1 mutation, FGFR2 fusion, and other tumor-agnostic targets such as BRAF V600E mutation, RET fusion, KRAS G12C mutation, NTRK fusion, dMMR, and TMB-high.^5,32 In other tumor types, the presence of certain mutations may be associated with reduced benefit from chemoimmunotherapy compared with targeted therapy,^36,37 suggesting the need to explore combination or sequencing strategies in BTC.

This study is limited by its retrospective, single-center design, which introduces the potential for selection bias in both treatment allocation and molecular testing. The modest sample size limits the statistical power for subgroup, biomarker, and multivariable Cox regression analyses. Incomplete and heterogeneity in NGS testing may have influenced alteration detection rates. Consequently, the findings should be interpreted as exploratory and hypothesis-generating rather than definitive. Despite these limitations, this study provides a pragmatic real-world benchmark that complements prospective trial evidence and highlights the importance of patient selection, toxicity monitoring, and genomic profiling in BTC management.

Conclusions

In a real-world setting, this study provides real-world evidence supporting the use of chemotherapy plus immunotherapy as first-line treatment for advanced BTC, with outcomes and safety broadly consistent with clinical trial findings. ECOG PS, hypoalbuminemia, and HER2 status emerged as key prognostic factors. These findings underscore the need for careful baseline patient assessment, proactive management of toxicity, and universal access to molecular profiling to optimize treatment selection. Future multicenter prospective studies should validate these results and explore personalized strategies for molecularly defined subgroups, particularly those with HER2-positive disease.

Supplemental Material

Supplemental Material - Effectiveness of Chemotherapy Plus Immunotherapy and Molecular Alterations in Advanced Biliary Tract Cancer: Real-World Evidence From a Single-Center Australian Cohort

Supplemental Material for Effectiveness of Chemotherapy Plus Immunotherapy and Molecular Alterations in Advanced Biliary Tract Cancer: Real-World Evidence From a Single-Center Australian Cohort by Jirapat Wonglhow, Hui-Li Wong, Michael Michael in Cancer Control

Footnotes

ORCID iD

Jirapat Wonglhow

Ethical Considerations

The study was conducted in accordance with the Declaration of Helsinki and approved by the Human Research Ethics and Governance of Peter MacCallum Cancer Centre, Australia (HREC: QA/121051/PMCC/September 3, 2025).

Consent to Participate

Patient consent was waived because of the retrospective nature of the study and approved by the Human Research Ethics and Governance of Peter MacCallum Cancer Centre. Patient information was anonymized for confidentiality.

Author Contributions

Conceptualization, J.W. and M.M.; methodology, J.W. and M.M.; software, J.W.; validation, J.W.; formal analysis, J.W; investigation, J.W.; resources, J.W., H.W., and M.M.; data curation, J.W.; writing—original draft preparation, J.W.; writing—review and editing, J.W., H.W., and M.M.; visualization, J.W.; supervision, J.W. and M.M.; project administration, J.W. All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.*

Supplemental Material

Supplemental material for this article is available online.

References

Nakanuma

Kakuda

. Pathologic classification of cholangiocarcinoma: new concepts. Best Pract Res Clin Gastroenterol. 2015;29:277-293. doi:10.1016/j.bpg.2015.02.006

Ilyas

Gores

. Pathogenesis, diagnosis, and management of cholangiocarcinoma. Gastroenterology. 2013;145:1215-1229. doi:10.1053/j.gastro.2013.10.013

Menon

Garikipati

Roy

. Cholangiocarcinoma. Treasure Island, FL, USA: StatPearls Publishing; 2025. https://www.ncbi.nlm.nih.gov/books/NBK560708/

Khan

Tavolari

Brandi

. Cholangiocarcinoma: epidemiology and risk factors. Liver Int. 2019;39(Suppl. 1):19-31. doi:10.1111/liv.14095

Vogel

Bridgewater

Edeline

, et al. Biliary tract cancer: ESMO clinical practice guideline for diagnosis, treatment and follow-up. Ann Oncol. 2023;34:127-140. doi:10.1016/j.annonc.2022.10.506

Valle

Wasan

Palmer

, et al. Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer. N Engl J Med. 2010;362:1273-1281. doi:10.1056/NEJMoa0908721

Bouattour

, et al. Durvalumab or placebo plus gemcitabine and cisplatin in participants with advanced biliary tract cancer (TOPAZ-1): updated overall survival from a randomised phase 3 study. Lancet Gastroenterol Hepatol. 2024;9:694-704. doi:10.1016/S2468-1253(24)00095-5

Kelley

Ueno

Yoo

, et al. Pembrolizumab in combination with gemcitabine and cisplatin compared with gemcitabine and cisplatin alone in patients with advanced biliary tract cancer (KEYNOTE-966): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2023;401:1853-1865. doi:10.1016/S0140-6736(23)00727-4

Testa

Pelosi

Castelli

. Cholangiocarcinoma: molecular abnormalities and cells of origin. Technol Cancer Res Treat. 2023;22:1533033822112868. doi:10.1177/15330338221128689

10.

Stenzinger

Vogel

Lehmann

, et al. Molecular profiling in cholangiocarcinoma: a practical guide to next-generation sequencing. Cancer Treat Rev. 2024;122:102649. doi:10.1016/j.ctrv.2023.102649

11.

Carotenuto

Sacco

Forgione

Normanno

. Genomic alterations in cholangiocarcinoma: clinical significance and relevance to therapy. Explor Target Antitumor Ther. 2022;3:200-223. doi:10.37349/etat.2022.00079

12.

Valéry

Vasseur

Fachinetti

, et al. Targetable molecular alterations in the treatment of biliary tract cancers: an overview of the available treatments. Cancers (Basel). 2023;15:4446. doi:10.3390/cancers15184446

13.

Abou-Alfa

Macarulla

Javle

, et al. Ivosidenib in IDH1-mutant, chemotherapy-refractory cholangiocarcinoma (ClarIDHy): a multicentre, randomised, double-blind, placebo-controlled, phase 3 study. Lancet Oncol. 2020;21:796-807. doi:10.1016/S1470-2045(20)30157-1

14.

Abou-Alfa

Sahai

Hollebecque

, et al. Pemigatinib for previously treated, locally advanced or metastatic cholangiocarcinoma: a multicentre, open-label, phase 2 study. Lancet Oncol. 2020;21:671-684. doi:10.1016/S1470-2045(20)30109-1

15.

Goyal

Meric-Bernstam

Hollebecque

, et al. Futibatinib for FGFR2-rearranged intrahepatic cholangiocarcinoma. N Engl J Med. 2023;388:228-239. doi:10.1056/NEJMoa2206834

16.

Javle

Borad

Azad

, et al. Pertuzumab and trastuzumab for HER2-positive, metastatic biliary tract cancer (MyPathway): a multicentre, open-label, phase 2a, multiple basket study. Lancet Oncol. 2021;22:1290-1300. doi:10.1016/S1470-2045(21)00336-3

17.

Subbiah

Lassen

Élez

, et al. Dabrafenib plus trametinib in patients with BRAF V600E-mutated biliary tract cancer (ROAR): a phase 2, open-label, single-arm, multicentre basket trial. Lancet Oncol. 2020;21:1234-1243. doi:10.1016/S1470-2045(20)30321-1

18.

Javle

Roychowdhury

Kelley

, et al. Infigratinib (BGJ398) in previously treated patients with advanced or metastatic cholangiocarcinoma with FGFR2 fusions or rearrangements: mature results from a multicentre, open-label, single-arm, phase 2 study. Lancet Gastroenterol Hepatol. 2021;6:803-815. doi:10.1016/S2468-1253(21)00196-5

19.

Zhu

Macarulla

Javle

, et al. Final overall survival efficacy results of ivosidenib for patients with advanced cholangiocarcinoma with IDH1 mutation: the phase 3 randomized clinical ClarIDHy trial. JAMA Oncol. 2021;7:1669-1677. doi:10.1001/jamaoncol.2021.3836

20.

von

Altman

Egger

, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147:573-577.

21.

Huang

W-K

Tang

Y-J

C-E

, et al. Real-world effectiveness and prognostic factors of durvalumab plus chemotherapy in a multicentric cohort with advanced biliary tract cancer. Oncologist. 2024;30:oyae306. doi:10.1093/oncolo/oyae306

22.

Dwivedi

AND

Jain

Dixit

. Gall bladder carcinoma: aggressive malignancy with protean loco-regional and distant spread. World J Clin Cases. 2015;3:231-244. doi:10.12998/wjcc.v3.i3.231

23.

Menon

Babiker

. Gallbladder Carcinoma. Treasure Island, FL, USA: StatPearls Publishing; 2025. https://www.ncbi.nlm.nih.gov/books/NBK442002/

24.

Dayyan

Zamani

Ajdarkosh

, et al. The prognostic role of albumin-bilirubin grade in the mortality of extrahepatic cholangiocarcinoma patients. BMC Gastroenterol. 2025;25:383. doi:10.1186/s12876-025-03979-9

25.

Liu

Wang

, et al. Association of albumin-bilirubin grade with survival outcomes in patients with cholangiocarcinoma. PLoS One. 2025;20:e0321758. doi:10.1371/journal.pone.0321758

26.

Seven

Karahan

Köş

Bayram

Sekmek

Esen

. The HALP score as a prognostic factor in metastatic biliary cancer. Clin Transl Oncol. 2025;27:1653-1657. doi:10.1007/s12094-024-03702-2

27.

Zeng

Wen

Wang

. Prognostic factors in gallbladder cancer: a comprehensive systematic review and meta-analysis. Hepatobiliary Surg Nutr. 2025;14:374-397. doi:10.21037/hbsn-23-502

28.

Ramirez

DER

Knøfler

Kirkegård

, et al. Prognosis related to treatment plan in patients with biliary tract cancer: a nationwide database study. Cancer Epidemiol. 2024;93:102688. doi:10.1016/j.canep.2024.102688

29.

Qin

, et al. Durvalumab plus chemotherapy in advanced biliary tract cancer: three-year overall survival update from the phase III TOPAZ-1 study. J Hepatol. 2025;83:S0168827825022019. doi:10.1016/j.jhep.2025.05.003

30.

Azizi

Lamarca

McNamara

Valle

. Chemotherapy for advanced gallbladder cancer (GBC): a systematic review and meta-analysis. Crit Rev Oncol Hematol. 2021;163:103328. doi:10.1016/j.critrevonc.2021.103328

31.

Azizi

Lamarca

McNamara

Valle

. Systematic review and meta-analysis of the efficacy of chemotherapeutic regimens in advanced gallbladder cancer: assessing current practice and treatment benefit. Ann Oncol. 2020;31(Suppl 4):S212. (Abstract). doi:10.1016/j.annonc.2020.04.457

32.

Kehmann

Jördens

Loosen

Luedde

Roderburg

Leyh

. Evolving therapeutic landscape of advanced biliary tract cancer: from chemotherapy to molecular targets. ESMO Open. 2024;9:103706. doi:10.1016/j.esmoop.2024.103706

33.

Fassan

Angerilli

Normanno

, et al. Practical guidelines for molecular testing of cholangiocarcinoma in clinical practice: Italian experts’ position paper. Crit Rev Oncol Hematol. 2024;194:104224. doi:10.1016/j.critrevonc.2023.104224

34.

Liang

Tang

Yang

. Clinical significance of HER2 overexpression in biliary tract carcinoma – a meta-analysis. Front Oncol. 2025;15:1534005. doi:10.3389/fonc.2025.1534005

35.

Janjigian

Kawazoe

Bai

, et al. Pembrolizumab plus trastuzumab and chemotherapy for HER2-positive gastric or gastro-oesophageal junction adenocarcinoma: interim analyses from the phase 3 KEYNOTE-811 randomised placebo-controlled trial. Lancet. 2023;402:2197-2208. doi:10.1016/S0140-6736(23)02033-0

36.

de Jong

Das

, et al. Novel targets, novel treatments: the changing landscape of non-small cell lung cancer. Cancers (Basel). 2023;15:2855. doi:10.3390/cancers15102855

37.

Kim

Halmos

. Choosing the best first-line therapy: NSCLC with no actionable oncogenic driver. Lung Cancer Manag. 2020;9:LMT36. doi:10.2217/lmt-2020-0003

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.62 MB